Greenberg’s Law of The Media

If a news item has a number in it, then it is probably misleading.

Want to end gun violence? End violent inequality

Greg Palast has the article Florida = Honduras: Inequality kills. Want to end the American shooting epidemic?.

The result of my scatter plot came as quite the surprise to me: there was just about no correlation between number of guns and number of gun homicides.”

In fact, “the correlation coefficient was -0.105871699.” That is, by a small amount, more guns meant fewer homicides.

So what DID prove a strong correlation? Homicides versus the “GINI” coefficient. GINI is the measure of income inequality in a nation.

The graph he presents makes little sense as an explanation of the excerpt above. I have yet to figure out exactly how the horizontal axis of the graph should be labeled to make sense of it. We must also remember that correlation does not prove causation. However, at least in this case the premise does make sense to me. Given this hint of what these measures might be indicating, I’d like to see someone publish a real study of the possibilities.

Reluctantly, I have to file this story in the category of Greenberg’s Law Of The Media – If a news item has a number in it, then it is probably misleading.

The tricks propagandists use to beat science

MIT Technology Review has the article The tricks propagandists use to beat science from January 22, 2018.

It’s a mildly interesting article, but I would be very, very wary of the suggested “solution”.

…the solution is clear: bigger, more highly powered studies. “Given some fixed financial resources, funding bodies should allocate those resources to a few very high-powered studies,” argue Weatherall and co, who go on to suggest that scientists should be given incentives for producing that kind of work. “For instance, scientists should be granted more credit for statistically stronger results—even in cases where they happen to be null.”

I have noticed this in the electrical engineering technical papers I have read over my 40 year career. One respected university had one bent in their work and a different respected university had a different bent. Knowing authors from both universities, I could tell which side of the discussion a paper would fall on based on which school the author came from. Faculty from both universities were the peers reviewing the papers published in peer reviewed journals. In this case, I don’t even think the bias was from the sponsor’s of the research because the companies I worked for sponsored research from both universities. Although I don’t doubt that there were influential engineers in the company that had received their advanced degrees from one university or the other.

Neither of the universities discussed above were MIT. However, I have had my dealings with sponsoring research at MIT, and I can tell you that the people there are human, too. To that, I guess I have to say #MeToo. I am aware that I have my own biases.

I have posted this article in the category of Greenberg’s Law of The Media – “If a news item has a number in it, then it is probably misleading.” This category applies to the subject of the article and to the article itself.

Flu Vaccine: Half a Statistic Is Worse Than None 1

NBC Nightly News had a story Growing concern over children dying of the flu.

There is one statement in the report that is a perfect example of how the media mislead you with half a statistic.

The report never told you what this number means. What did they expect you to learn from this? I can think of three possible conclusions you could take depending on what is the value of the statistic they did not report. What they failed to report was what percentage of the children who survived were never vaccinated.

In the figures below I have chosen three possible values for the missing statistics of the percentage of children who were not vaccinated that survived. Above each graph, I have put a label of what you might be able to conclude given any one of the green bars compared to the red bar.

In the above figure, of the children who survived they had a lower percentage of not being vaccinated so you might conclude there was an advantage to being vaccinated.

In the above figure, of the children who survived they had the same percentage of not being vaccinated so you might conclude there was neither an advantage nor a disadvantage to being vaccinated.

In the above figure, of the children who survived they had a higher percentage of not being vaccinated so you might conclude there was a disadvantage to being vaccinated.

Without seeing a green bar, there is nothing you can conclude from seeing the red bar alone. You might have concluded that certainly the vaccine had whatever advantage or disadvantage you had assumed before seeing the number. In other words, this half statistic may have made you more sure of the wrong thing.

A Revolutionary New Type of Lens Focuses All The Colours of The Rainbow Into a Single Point

Science Alert has the article A Revolutionary New Type of Lens Focuses All The Colours of The Rainbow Into a Single Point.

A brand new type of lens called a metalens has just passed a major hurdle. A metalens is a flat surface that use nanostructures to focus light, and it could change optics forever by replacing the traditional bulky, curved lenses we know.

This was an interesting read, but the following stopped me in my tracks.

Making a metalens like this is so tricky because different wavelengths of light move through materials at different speeds. That leads to focusing errors known as chromatic aberrations, which traditional lenses get around through curved surfaces.

And here I thought that lenses used curved surfaces as a way of changing the magnification of an image. In other words the curved surface is what makes it a lens. I went to WikiPedia to see what it had to say in the article Chromatic aberration.

There exists a point called the circle of least confusion, where chromatic aberration can be minimized.[6] It can be further minimized by using an achromatic lens or achromat, in which materials with differing dispersion are assembled together to form a compound lens.

This is the explanation I imagined. Of course, the Wikipedia has much more detail and talks about other techniques of correcting chromatic aberration.

I just think this is an example of what happens when an author with a tenuous understanding of a scientific topic tries to simplify an article to explain science to other people with a tenuous understanding of the topic. Don’t treat as gospel what you read in a news medium that has the word “science” in its name. Actually, such a medium may be no more trustworthy than a medium that makes no claim to be about science.

Here’s Why the Latest Conservative Talking Point About the $15 Minimum Wage Is Meaningless (and Here’s What You Can Do About It)

Civic Skunkworks has the article Here’s Why the Latest Conservative Talking Point About the $15 Minimum Wage Is Meaningless (and Here’s What You Can Do About It) – August 12, 2015 11:53 am

“It’s a sad fact that bullshit travels faster than truth on the internet, but if we all work together, we can stop the conservative agenda from drowning out our good news with their meaningless noise.”

There are two follow-up posts as of this writing.

Fox Business Twists the Truth Seven Times in 100 Seconds on Seattle’s Minimum Wage – August 13, 2015 12:52 pm

The Data on That Minimum Wage Report Continues to Fall to Pieces – August 14, 2015 10:54 am

This is a good series of articles to remember. Put it in your arsenal next time someone asks you to just tell them when did Faux Noise ever lie. You can also use it to cast aspersions on the “scholarship” of The American Enterprise Institute. It isn’t scholarship unless they are trying to teach how to lie with statistics.

Here are some laws to remember.

Greenberg’s Law of Counterproductive Behavior

If you see a behavior that seems to you to be counterproductive, perhaps you have misunderstood what the actor’s real goal was.

Greenberg’s Law of The Media

If a news item has a number in it, then it is probably misleading

Study: Social Security in REALLY bad shape

USA Today has the story Study: Social Security in REALLY bad shape.

“The projections developed by the Office of the Chief Actuary for the Trustees Reports are intended to reflect all aspects of future possible trends in demographic, economic, and programmatic factors, given current Social Security law,” Goss and other SSA officials wrote. King and Soneji’s projections “were within the range of reasonable uncertainty as specified in the Trustees Report, and therefore should cause no alarm.”
“Fair, transparent and accurate forecasts give Congress more of a chance to consider of all the policy proposals to preserve the solvency of Social Security,” King said. “And it’s easier to make changes to Social Security now than in the future.”

The other unmentioned assumption has to do with the wealth and income distribution. Since income has been shifted from the middle class to the wealthy in the last 30 years or so, the level of social security contributions from the middle class has declined below previous expectations.

If the SSA actuaries disclosed the impact of income distribution on their calculations, then I would be all for increased transparency. The Republicans would ignore that issue, but one can only hope that at least one politician who was looking out for the middle class would keep harping on it. Who is going to be the politician to do it after Bernie Sanders finishes his term as President?

I have put this story in the category of Greenberg’s Law of the Media.

If a news item has a number in it, then it is probably misleading.

In this case, there are several things that are misleading.

  • The numbers they give you are intended to lead you to one of the solutions. If they had put in the numbers that they left out, it might lead you to think of other solutions.
  • When the article talk about using fiscal gap accounting methods they say, “Under this accounting system, SSA’s projected unfunded liabilities would be $24.9 trillion (instead of the $10.6 trillion projected in 2088).” They don’t explain that you cannot draw the same type of conclusions from a larger number calculated by a different method, than you would if that same larger number had been calculated by the traditional method. In fact the implication is that the larger number is more “truthful” in a sense. It makes no sense to imply that.

If I didn’t make clear the reasoning behind my judgment, tell me why you think I am wrong, and I will try to tell you what I left out of my explanation. I can’t guess all the things that were going through your mind when you read this compared to all the things that were going through my mind when I wrote it.

Ted Cruz says satellite data show the globe isn’t warming. This satellite scientist feels otherwise

The Washington Post has the article Ted Cruz says satellite data show the globe isn’t warming. This satellite scientist feels otherwise by Chris Mooney.

Mooney discusses the data that Cruz uses to come to the conclusions that he does.

To explore Mears’s views further, I did one thing journalists can do when covering the climate views of presidential candidates — I contacted the researcher. And his response was quite critical of Cruz’s approach to the evidence on this issue:

Mr. Cruz (and others who seek to minimize the threat posed by climate change) likes to cite statistics about the last 17 years because 17 years ago, the Earth was experiencing a large ENSO [El Nino-Southern Oscillation] event and the observed temperatures were substantially above normal, and above any long-term trend line a reasonable person would draw. When one starts their analysis on an extraordinarily warm year, the resulting trend is below the true long term trend. It’s like a pro baseball player deciding he’s having a batting slump three weeks after a game when he hit three homers because he’s only considering those three weeks instead of the whole season.

So if you have been tempted to fall for this meme that there hasn’t been global warning recently, remember this is the trick that has been played to come to this conclusion.

They could have mentioned the President Reagan fans measuring the performance of the economy from the depths of the Reagan induced depression to the end of his term to convince you that Reagan was wonderful on the economy.

I have categorized this post as an example of Greenberg’s Law of The Media – “If a news item has a number in it, then it is probably misleading”. Actually it is probably a counter example. In this case the news article debunks one of the common techniques of misleading.

Boston scientists say triglycerides play key role in heart health

The Boston Globe has the article Boston scientists say triglycerides play key role in heart health.

A massive genetic study led by a Boston cardiologist has identified a subset of people who carry rare mutations that cause them to have dramatically lower levels of triglycerides in their blood. Those people, in turn, were 40 percent less likely to have heart disease than people who didn’t have the mutation.
The findings suggest that scientists should be looking for a way to mimic what the body does in those people with naturally low levels of triglycerides.

I posted my comments to the article on the newspaper’s web site.

Another example of the innumeracy of the press. By now there are hundreds of thousands of statisticians crying out “Correlation does not mean causation.”

It could very well be that the mutation’s side effect of lowering triglycerides may have nothing to do with causing a lower heart disease rate. It might also be that attempting to lower triglycerides by artificial means will have damaging unintended consequences. What countervailing mechanisms will the normal human body bring into play as a consequence of an artificial lowering of triglycerides? It may be that such a lowering without the gene mutation could be fatal.

Is anybody asking these obvious questions? If the doctors involved in this study aren’t aware enough to ask these questions, maybe it is too much to expect the “medical experts” in the news media to think of asking these questions.  Maybe it takes a person like myself with no degree in anything medical to see the forest among the trees.

It may be time for everybody to take another look at RichardH’s post on this blog Diversion–Highway Fatalities and Lemons.

Mindless Budget Reporting: Fooling Some of the People All of the Time

The PBS story,  Mindless Budget Reporting: Fooling Some of the People All of the Time by Dean Baker talks about an example of Greenberg’s Law of The Media. Baker is castigating a report in The New York Times.

“A plan by House leaders to cut $40 billion from the food stamp program — twice the amount of cuts proposed in a House bill that failed in June — threatens to derail efforts by the House and Senate to work together to complete a farm bill before agriculture programs expire on Sept. 30.”

The problem with this description of the Republican plan is that the proposed cut of $40 billion is supposed to be over a 10-year budget window, not a single year. (The Republicans want to cut the food stamp budget by 5 percent, not 50 percent.) This information is not reported anywhere in the article. As a result, even a very intelligent and extremely knowledgeable person like Krugman could read through the piece and be off by a factor of 10 in his understanding of the size of the proposed cuts.

One of the ways Greenberg’s Law is demonstrated is to give us a number out of context. You are obviously supposed to infer that the number illustrates some point that the reporter is implying, but you are never given the context to judge whether the desired inference is correct.  It is unlikely that the reporter knows whether the desired inference is correct.

Baker is correct that all you know is that it is a large number.  If you don’t know whether it is over 1 year or 10, or what fraction it is of the budget, or how this government spending compares to the spending of the corporate sector under similar circumstances, then you have no idea if the number is too large, too small, or just about right.  However, your thinking about the matter has been prejudiced by the report.  Because of this, you might come away from reading or hearing the story with less knowledge than you started with.

Sort of like all of Faux Noise, the more you watch, the less you know.

Charter schools in Boston score higher on key tests

The Boston Globe has the article Charter schools in Boston score higher on key tests.  If you are not a Globe subscriber, the only text you get to see is:

Boston charter schools outperform other public schools on three popular barometers of achievement — the MCAS, the SAT, and the Advanced Placement exams — but tend to have lower four-year graduation rates, according to a study being released Wednesday.

If you read the newspaper or have full access to the site, you will see the following about half way through the article:

In Boston, there are 25 charter schools.

The study examined 3,400 students who sought admission to one of the six charter high schools in Boston between fall 2002 and 2008. (The study excluded two charter high schools that closed during that period because of low performance.)

I commented on the article which reported on a study done at MIT.

If there had been more room, the headline might have said “Charter schools score higher on key tests except for the ones that don’t”  It is convenient how two schools that would have lowered the averages for the Charter schools were taken out of the study.  Perhaps the people conducting the study and doing the statistical analysis could have excluded a similar proportion of low performing public schools from the study.

With MIT accepting huge amounts of funds to build buildings named after the infamous Koch brothers and then this story, perhaps it is  true that MIT is selling its soul to the devil in order to raise funds.  Now when MIT calls me for an alumnus donation, I just tell them to put it on the Koch brothers’ tab.